Muneeb.

Essay 04 · 09 Mar 2026 · 5 min read

Optimize Revenue, Not Clicks

Every metric in an ad account is a proxy. The question is how many steps sit between the proxy and the money.

Fig. 01 · The narrowingVolume in, less volume out, and where it happens

Click-through rate is a proxy for interest. Cost per lead is a proxy for efficiency. Both are several steps removed from revenue, and every step is a place where the relationship can break.

This is not an argument for ignoring them. They are fast, and revenue is slow. It is an argument for knowing exactly how far from the money each one sits.

How the break happens

A campaign optimizing for cost per lead will find you cheap leads. It is very good at this. It will find the audiences that fill in forms, which is a different population from the audiences that buy.

Six weeks later cost per lead is down 30% and pipeline is flat, and the account looks like it is working. Nothing malfunctioned. The system optimized precisely for what it was asked to optimize for.

Cost per lead down 30%, pipeline flat. Nothing malfunctioned — the system optimized for exactly what it was asked to.

Move the target down the funnel

The fix is not a better proxy. It is a proxy closer to the money, fed back to the platform.

  • 01Send qualified leads back, not raw leads. Even a crude qualification flag changes what the algorithm chases.
  • 02Where volume allows, send closed-won and deal value. Sparse but honest beats dense and misleading.
  • 03Weight by segment when data is thin. If enterprise deals are worth 8x, say so, rather than waiting for enough conversions to prove it.

There is a real trade-off. Optimizing on a downstream event with a long sales cycle means slower learning and a noisier signal. In a business with a ninety-day cycle you cannot wait for closed-won to steer daily decisions.

So you use both. The fast proxy to steer the week, the slow number to check the quarter — and when they disagree, the slow one wins.

The uncomfortable version

Optimizing for revenue usually makes the dashboard look worse. Cost per lead goes up. Volume goes down. The chart that used to be green is no longer green.

That conversation is easier to have before you change the target than after. Agree with whoever reads the dashboard on what success looks like, and get it in writing, because in eight weeks somebody is going to ask why leads are down.

Syed Muneeb Rehaman

AI-first Growth Operator · 09 Mar 2026

SM